A platform that launched in July with an AI assistant called Sarah has now put its tools into two mobile applications, one for licensed agents and one for buyers, reshaping how listings, valuations and follow-up are handled in the field.
Per an 11 August 2026 announcement from Ghostly Labs, the Phantom Realty Engine has rolled out a pair of apps that surface the same MLS-driven intelligence to both sides of a transaction.
Key facts:
- Ghostly Labs released the agent and client apps on 11 August 2026.
- The platform's assistant, called Sarah, reads each client's live MLS feed, not a national aggregate.
- The platform is active across Canada and 30 US markets.
- In the platform's July launch, Sarah handled more than 15,000 cold leads in its first deployment.
- Per the release, the platform warmed nearly half of those leads and converted more than 1,400 into active buyers.
- Average user time on the platform rose 6.4 times above the prior baseline.
What the two apps actually do
The buyer-facing app gives each agent's clients a white-labelled product. Buyers can scroll through every listing in their agent's market, save the ones they want and bring a spouse or partner into the same shortlist. They can chat with Sarah about a specific home or neighbourhood and get answers pulled from their agent's live MLS data at any hour. They can request a valuation of their own property, which the platform routes back to the agent as a seller lead, and they can book a showing straight from the app.
The agent app puts the platform's CRM in a phone. From the field, an agent can pull a full comparative market analysis, run comparable sales, check price per square foot, track days on market and see what is actively competing for the same buyer. They can search the MLS, push listings to clients, refresh saved searches and watch live activity on their site. Sarah also flags which contacts have gone cold, who is warming up and drafts follow-up messages in a tone matched to how each contact tends to communicate.
Why the MLS-first architecture matters
Per the release, the platform sources data from each client's own live MLS rather than a national rollup. That structural choice is the real story. Neighbourhood-level questions get neighbourhood-level data, not a national average dressed up as a local one. For an industry that has been burned by generic AI tools that confuse Phoenix with Phoenixville, that pitch lands.
Our analysis
The launch numbers are loud but secondary. The interesting question is whether MLS-first AI becomes the default architecture for agent-facing tools in 2026 and beyond. Per the release, Sarah worked through more than 15,000 cold leads in her first deployment, warmed close to half and converted more than 1,400 into active buyers. That is a conversion curve most agent CRMs cannot match. Average user time on the platform rose 6.4 times the prior baseline, which suggests the AI is not just answering questions but keeping people on the surface.
For brokerage owners, the practical question is integration. If buyers get their own branded app, the brokerage brand becomes a distributed asset, sitting on every client phone rather than living behind a single website. That is a different customer-acquisition posture. Rob Deaton, a licensed Realtor and co-founder of the platform, said in the release that previous platforms handed him a website and stopped there, his buyers drifted to other sites and got resold back to him as leads he had already paid for. Josh Shampo, founder of Ghostly Labs, said in the release that the apps were not a phone version of the website but the same intelligence and live data on whatever screen a person happens to be holding.
A worked example
Say you run a 60-agent brokerage in a mid-sized Canadian city. You adopt the platform across all 60 agents. Each agent gets the agent app with Sarah doing CMA and follow-up. Each agent's buyers get the client app branded to the brokerage. Your buyer pipeline shifts from email and SMS into a branded app you control. Your listing inventory sourced to the platform comes from buyers requesting valuations of their own homes, which feed seller leads back to your agents. Net effect: more buyer-side engagement, more seller leads, lower lead-cost per closed transaction.
Who should care
| Role | Why this matters |
|---|---|
| Brokerage owner | Distributed brand on every client phone; potential to lower cost per closed deal. |
| Independent agent | CMS, CMA and follow-up drafted from the field, not from a desk. |
| Team lead | Lead-routing and quiet-contact flagging built into the agent app. |
| Investor | Watch whether the conversion curve holds as the user base scales. |
What to watch next
Watch coverage expansion into additional US markets. Watch MLS integration partnerships and whether incumbent CRMs respond with native AI features or partnerships of their own. Watch the conversion data: if the 1,400-of-15,000 ratio holds at scale, the unit economics of an agent shift materially.
Key takeaways
- The platform reads from each client's live MLS feed, not a national aggregate.
- Sarah turned more than 15,000 cold leads into over 1,400 active buyers in the platform's first deployment.
- The dual-app structure gives brokerages a distributed customer-acquisition asset.
- Coverage is currently Canada plus 30 US markets.
- Average user time on the platform rose 6.4 times versus the prior baseline.
Frequently asked questions
What did Ghostly Labs launch?
Two mobile applications for Phantom Realty Engine: an agent app for CRM, CMA and MLS search, and a client app for buyers to swipe listings and ask Sarah questions.
How is this different from existing AI tools?
Per the release, Sarah reads from each client's live MLS feed, not a national aggregate, so neighbourhood-level questions get neighbourhood-level data.
Where is the platform available?
Across Canada and 30 US markets as of the launch.
What results did the platform report in its first deployment?
Per the release, Sarah handled more than 15,000 cold leads, warmed nearly half and converted more than 1,400 into active buyers.
Why does the client app matter for brokerages?
Each buyer's app is branded to the agent or brokerage, giving the firm a distributed customer-acquisition asset rather than a single website.